Improvement of a Deep Learning Algorithm for Total Electron Content Maps: Image Completion

Observations with a global coverage are very important for space physics research and space weather monitoring. However, due to the technical limitations, it would be very expensive or even impossible to achieve a seamless global coverage even with advanced observational devices. It would be useful to fill missing data gaps to create a global map from the available data, but up until now this has been very challenging. Fortunately, the deep learning method, a recent breakthrough in artificial intelligence, may provide an effective way to solve this problem by making full use of data from reliable observations. In this paper, a promising deep learning algorithm, deep convolutional generative adversarial network (DCGAN), is investigated to fill the missing data of total electron content (TEC) map images. The direct use of DCGAN fails to fill missing data for the completion of TEC maps because there are always missing TEC data in some regions, such as oceans, where the features vary with time and geophysical conditions. Thus, no useful information can be utilized by DCGAN to achieve a meaningful image completion. In order to overcome this shortcoming of the original DCGAN method, a novel regularized DCGAN (R-DCGAN) is proposed by adding an extra discriminator and some widely used reference TEC maps from the International Global Navigation Satellite Systems Service Ionosphere Working Group. The proposed R-DCGAN method generates satisfactory ionospheric peak structures at different times and geomagnetic conditions, which demonstrate its effectiveness on filling the missing data in TEC maps. The proposed R-DCGAN framework can be readily extended to a broad application in other fields of space sciences, particularly for addressing the missing observation data issues. Plain Language summary This paper proposes an improved deep learning algorithm, regularized deep convolutional generative adversarial network (R-DCGAN), for the image completion of total electron content (TEC) maps. The traditional DCGAN (which is a very popular and powerful deep learning algorithm for image completion, such as human face images) needs the training data to be completed observations. Since there is lack of distinct features in the missing data part of the training data that can be utilized by DCGAN to fill these missing values, DCGAN fails to directly learn the observation with data missing. In order to overcome the shortcoming of the original DCGAN method, an improved algorithm, R-DCGAN, is proposed to fulfill missing data completion for the Massachusetts Institute of Technology-TEC maps. The R-DCGAN is designed from DCGAN, with an extra discriminator and the reference TECs. The R-DCGAN produces satisfactory ionospheric peak structures at different times and geomagnetic conditions, and the results demonstrate that the deep learning algorithm is promising to fill the missing data.